I load 8 16-bit integers into a 128 bit variable. I compare each 16 bit quantity against a scalar. The result is FFFF or 0000 for each of the 16 bit quantities. How do I extract the low 8 bits of each quantity and write them to memory?
Here is the C scalar code. You are welcome to just use the intrinsic functions available in the VS C++ compiler. I am writing the results to an array of uchars.
short v = wfin[i];
predout[i] = (v < tupper && v > tlower) ? 1 : 0;